Earthworks in Joydens Wood, Bexley, Kent

( io ) EARTHWORKS LN JOYDENS WOOD, BEXLEY, KENT. BY A. H. A. HOGG. JOYDENS WOOD lies about 1£ miles to the south-east of Bexley, to the east of the river Cray, on a hill of sandy gravel rising to about 250 feet above Ordnance Datum. It has long been known to contain many deneholes associated with banks and ditches. Most of these seem to be the remains of a farm site, probably medieval, and its associated fields, lying at the junction of two ancient tracks. To the west there is a dyke, probably of earlier date, facing towards the Cray valley. Little excavation has been done, and relics are scarce, but in spite of this the remains are of interest. The farm site is unusually complete and well preserved owing to the absence of later disturbance, and provides a useful indication of the character of mediaeval agriculture in this area. It would be unwise to generalize without further work on other sites, but it may be noted that strip lynchets, generally regarded as typical of the early mediaeval period in most of southern England, are decidedly rare in Kent, and it seems possible that the Joydens Wood type of site may have been more usual. The dyke belongs to the group of " grimsditches " partly surrounding London, and is the only one in which any excavation has been carried out. It is also mentioned in a ninth century boundary survey which is itself of interest and which has not previously been worked out in detail. Only two accounts of the site which are of any value have been published, and neither deals fully with the remains. F. C. J. Spurrell has published a plan1 on a scale of approximately 6 inches to one mile, together with a short description, in his paper on " Deneholes." The plan contains a few minor errors, but is generally very acourate and complete. T • * * *' / V * 'o V-. r V?Wv STANKEV WOOD \ v - ^ % 4L-+ * CAVtV &PKINC ,•:•'•' "v7 7p?— -"••••• * i $ '•.DCMCHOLCX" . r . £ t m „ „ M • CLUSTER •*. .-3 ' *- F • * - • : » Z • ; r« N •*v... -A \ ^Tfifr. 4 nouNP , \ \ ^ r^ v\ r , : . » i * ^ » » , > * '*• ,:";'. *: v"* ( .^/^ ; *kD &»AVC • S*L°°-' J*& #**> »L*NT«T10N • i * * /? •N.. :' • - " . • « - •'-- ^••^T--~ * *HABlOW VCI.U /••-. / ) ( >>( \ ••:. c: £ / A • .-- \ f / . I K / \ N ••-•" •-.1-VN.trteTS ' \ -."" / j 1 \ \ HUT SITEfC?- \ \ i \ r ^ \ ? ^-^ f^Ki \y^ w >.•S* '*»&.<: S^^S*"*" C T " « W « I I I I » " 3««¥(Ul o? af, •V # ' \ RESURVEYED. FCJ.S. SANK DITCH J;;B» CONTOURS " PARISH BOUNDARY EDGE OF WOOD •""' MODERN TRACKS DENEHOLE o SECTION — PROFILE 2 «" EARTHWORKS IN JOYDENS WOOD NEAR BEXLEY. KENT VIII &IX. IOPO ISOOFEET T6o METRES Fio. 1. EARTHWORKS IN JOYDENS WOOD. 11 All subsequent general plans seem to have been based on this survey, and without it the writer's re-survey would have been impossible. The results of excavations made in a square •earthwork in the eastern part of the wood during 1925 by Mr. F. C. Elliston Erwood, F.S.A., have been described by him in a paper on " The Roman Antiquities of North-West Kent,"2 accompanied by a new plan of the square earthwork, and a small scale plan of the other banks, showing a few in addition to Spurrell's. These papers are referred to as S and E in the following account. In addition to the above, a plan of the square earthwork, with profiles, and an enlarged plan of the other banks based on S were published by Colonel 0. E. Ruck in 19063 and there have been a few other minor references.4 From 1924 onwards the eastern part of the wood has "been sold as small holdings and building plots, but although much cut up and rendered difficult of access very few of the banks have been destroyed. During 1934 the writer re-surveyed the site on a scale of 1/2500, checking S, both on the ground and from two sets of aerial photographs.5 On the plan (Fig. 1, reproduced to 1/7500 scale) those banks which could not be examined are indicated by a different convention. Deneholes, except where they occur on lines of bankwhich have been re-surveyed are based either on the 1/2500 O.S. map or on S. No attempt lias been made to plan the two denehole clusters, as the exact distribution did not seem of sufficient importance to justify the labour involved and a check in Cavey Spring showed both S and the O.S. map to be incorrect. Stankey Wood is now built over. It is clear that the earthworks belong to several periods, but since these are not always certainly distinguishable it was decided to prepare a plan showing all banks of whatever age, and to supplement it by a diagrammatic period plan. (Kg. 2.) The remains will be described in the order in which they seem to have been constructed, with the exception of the dyke, which cannot be treated without considering other examples of similar earthworks. 12 EARTHWORKS IN JOYDENS WOOD, /*MOUND Ya W W ' DAM Ye ••• OOO FEET -_- TRACK WAYS© "~ DYKE(E) — PERIOD ttC — PERIOD ET MISCELLANEOUS • DENEHOLES EARTHWORKS IN JOYDENS WOOD. DIAGRAMMATIC PLAN. Era. 2. BEXLEY, KENT. 13 (I.) The earliest remains on the site seem to be the two tracks AB, CD, which cross at the square earthwork. There is no certain evidence of the date of the dyke EFGH relative to these tracks, but it seems reasonable to suppose that their existence formed one of the principal reasons influencing the position of the dyke. The track AB generally forms a slight hollow between two low banks (CD vert. 12-18 ins.), and seems to have sent off a branch towards Dartford. CD, except where altered by later banks, forms a very faint intermittent hollow track. It may be noted that it leads towards the ford at Dartford, and is unlikely, therefore, to be earlier than that crossing. Its extension to the south is uncertain. (II.) The dyke seems to have been the first earthwork to be constructed, but will be discussed later. (III.) It seems to have been followed by a system of agriculture involving the cultivation of small fields and the use of chalk to fertilize the sandy soil. The remains of this period fall into two groups, J and K, both characterized by small banks (CD 12-24 ins.), sometimes forming lynchets, with deneholes on the line of the earthworks. It seems probable that it was during this period that the deneholes were dug on the line of the dyke at FG, and the dyke there was converted into a lynchet. The most reasonable explanation of the deneholes is that they provided the chalk to help the poor lime-free sandy soil. The pits at m seem to be neither deneholes nor pit dwellings, but are more probably sandpits, similar to those in Rose Wood, near Ightham.6 (IV.) In the last phase of the occupation of the site the double square earthwork and the banks (CD ca. 3 ft.) surrounding the three adjacent fields were constructed. These banks are later than the track CD, and also overlie, and in some cases make use of, the banks of group J. The banks UTF on the west of the old track CD are shown on the Tithe Award map as the boundary between Baldwynes Wood and Joydens Wood. On the west of the site these latest banks cross the small bank LM, which itself is later than the dyke. LM seems to belong to period III, but the bank shown connecting LM to group K is very faint indeed, and it would be unsafe to base any argument upon it. 14 EARTHWORKS IN JOYDENS WOOD, It is possible that the small inner square NPQR maybe earlier than the other banks but this is unlikely, for the corner N is very large, owing to the extra earth which comes from the ditch at a right-angled corner, whereas Q shows no increase in size. P and R have both been disturbed. If the original work had been NPQR, it seems probable that all the corners would be equally enlarged. It is almost, certain therefore that NPQR is contemporary with the banks which touch it. The other banks of this group are certainlyall of one period. It is clear from the layout of the enclosures on either side of the track CD that the large enclosures STUV, XYZ are contemporary, but at first it appears that, the double square may have been constructed later. It will be shown below that this cannot be the case. An examination of the junction T shows that the three banks meeting there are all of one build. At S the bank ST crosses the. ditch of the outer square, which cannot therefore be the more recent. But at V the ditch of the outer square cuts the bank UV which is contemporary with ST, so the outer square is not earlier than ST and must, therefore, be contemporary with it as it has been shown that it cannot be later. It is suggested in E that WW formed the NE side of the outer square, but there is no bank on the line WV to. complete the other side, and WW has more resemblance to> the banks of period III. A plan of the square earthwork (Fig. 3) is given on a. larger scale and shows the arrangements of the internal banks (CD ca. 2'), and the position of the foundations found. The small square foundation dry built of flints is that discovered by Mr. Elliston Erwood. Fifty pieces of medieval potterywere found near it. The other walls, further to the south, are exposed in the sides of a small rubbish pit. They are dry built of chalk and flint, about 18 inches high, and are accompanied by a layer of thin roofing tiles 18 inches below the present surface, evidently from the fallen roof. No complete tile could be recovered, but they are certainly post-Roman. The absence of scattered building material suggests that the superstructure was of timber. The ruined building seems to have been buried by the construction of BEXLEY, KENT. 15 the rectangular mound in which it stands, but this is not certain. Sections cut by Mr. Elliston Erwood in the other banks 4 \ ^' DENEHOLE ''•"": :":"""","i«:\v.v»v^*w r*Iif f'.:::;!.': \11 I t 11i ./•.'.SSiViW ;i!!».U:'.5'. 111 j j 11 rj j t, DENEHOLE?! Uiiiii't'iiViiV: „s WELLo 11. i' V.'' ' '%'_WALLS|?>, i;5;»ifMJJff; i i i i ! ( ( ( : ! ! . ' ! i EARTHWORK, JOYDENS 5 0 0 FEET. • i • > i IOO METRES. '!??! i * i ! ] ~ \ i :0 ; »t r..:;:J::; :i!'•'.:: :.r. WOOD. FIG. 3. gave no information as to'their structure (owing to the way in which the sandy soil " heals ", leaving no trace of disturbance), but they are far too regular to be dismissed as natural. The low lynchets in the north corner indicate that part of the area was cultivated. There is a deep well in the 16 EARTHWORKS IN JOYDENS WOOD, eastern part of the earthwork, steyned with flints. It is doubtful whether the pit at the east corner is in fact a collapsed denehole, as the bottom is flat, not bowl-shaped. It is more likely to have been a small sandpit. Spurrell records the finding of Roman pottery within the earthwork,7 but nothing of Roman date has since been found on the site, and it seems probable that the pottery was coarse medieval ware, which could formerly be picked up on the surface. Apart from this doubtful record, there is nothing to suggest a Roman date for the earthwork. The relics found have all been of medieval date, and, apart from any other point, the lack of entrances is not characteristic of Roman work. The plan can be very closely paralleled by the site of Preston, Surrey, first mentioned about the twelfth century.8 It is practically certain, therefore, that the earthwork in Joydens Wood is post-Roman. The mediaeval foundations of chalk and flint suggest very strongly that some of the deneholes belong to approximately the same date. Lesnes Abbey is known to have held two manors, Baldwyns and Ocholt,9 in this district. The name of Baldwyns has survived, and in modern times the house stood on the site now occupied by the Mental Hospital north of the wood. As mentioned above, some of the banks of Group IV form the south boundary of " Baldwynes Woods " in Dartford Parish Tithe Award Map. The name of Ocholt, however, has entirely vanished, but its position agrees very closely with that of the earthworks of Group IV. It is known to have stood on the boundary of Dartford and Wilmington parishes, and part of Ocholt lay between the land of Simon le Hert and " Le Haec ". These names may survive in Lower and Old Harts (adjacent field names on the T.A. map about 1,500 feet south-east of the square earthwork), and possibly in Hook Green. On the Wilmington T.A. map the woodland east of the track CD is described as Hazely, Baldwyns, and Harts Woods, but no boundaries within the wood are shown. The T.A. maps also show that most of the wood north of the square earthwork was open at that date. Near the square, the bank WW formed the boundary of the wood. BEXLEY, KENT. 17 Precise identification is not possible, but it seems extremely probable that the square earthwork and attached fields are the remains of Ocholt manor or in some way connected with it. The disappearance of the name may be due either to its absorption by Baldwyns manor, or the two manors may be identical. Before considering the dyke, some of the other remains in the wood may be mentioned, but the small banks which were probably connected in some way with the planting of the woods, and the occasional hollow trails and gravel pits, do not need any further description than that given on the plan. Va. The mound near the north corner of the square earthwork is now practically destroyed. It was excavated by Spurrell, and nothing was found. It seems to have stood within a small uncultivated rectangle, surrounded by slight lynchets resembling the Group III banks. It may have formed the site of a wmdmill, but there is no evidence. Vb. A group of lynchets on the south side of the valley. Vc. Corner of an earthwork (OH 50', CD vert. 2-5') on the top of a small hill forming the end of a promontory. Most of the site has been destroyed by old gravel pits. Vd. Two probable hut sites and a round pit, date unknown. Ve. A small dam across the head of a valley. Vf (not on diagram). Hadlow Well, a spring pond in the west of the wood on the parish boundary. At present the only permanent source of water in the wood. The dyke, II, perhaps the most interestmg of the earthworks in the wood, has been fully described in a note in Antiquity.10 It is to be identified with a " faestendic " referred to in a Saxon boundary survey dated A.D. 814. This survey is discussed in Appendix I. The description which follows is substantially that given in Antiquity with a few additions. The plate of profiles is not reproduced here, but the points at which they were taken are indicated. The northern end of the dyke lies on the east side of 5 18 EARTHWORKS IN JOYDENS WOOD, a shallow valley running northwards, but the steep slope on which the dyke ends dies out a short distance to the south. It continues to the north, however, forming a natural extension of the dyke almost to Crayford. At Tile Kiln Lane, 1,400 feet south of the north end, the dyke curves west to cross the valley. For the next 800 feet, between the lane and the wood, cultivation has destroyed the dyke, but the line is continued by a lynchet facing east, just below the brow of the hill on the west side of the valley. On entering the wood, the dyke curves through a right angle (Profile E), and runs west for 600 feet, a footpath following the top of the bank. At a collapsed denehole on the north of the path the dyke turns sharply at right angles towards the south. This is the " Bend " of the charter boundaries. South of the footpath the dyke has been almost destroyed for 150 feet. At the end of this stretch is a gap (1) where the track AB crossed the dyke. It seems to be original, but is too much obscured for certainty to be possible. The rest of the dyke is fairly uniform in profile, but from 500 to 1,100 feet from the path it seems to have been enlarged (Profile D). A section (1) was cut at a point 600 feet from the path, and is described below. Near this point the dyke is crossed by the small bank LM. The dyke continues to a point 1,400 feet south of the footpath, where it dies out on the edge of a steep-sided valley. This stretch of dyke faces slightly downhill. After about 700 feet the dyke reappears on the other side of the valley. In the gap (2) are traces of cultivation, and a slight bank and ditch, smaller than the dyke (OH30', CD vert. 3' approximately), but facing the same way. A hollow trail follows the ditch and continues to the top of the hill. After its reappearance the dyke runs for 1,500 feet just below the brow of the hill, facing uphill. A cart track follows the ditch for 1,100 feet. 1,200 feet from the point at which the dyke leaves the south side of the wood is a gap (3) 50 feet wide and 4 feet deep which carries the surface drainage from outside the dyke into the head of the valley. It seems to be original. 500 feet further south is another BEXLEY, KENT. 19 break (gap 4) where the line of the dyke crosses a patch of marshy ground. This also seems to be origmal. The dyke extended to the south of the wood but the ground has been cultivated for many years. Two sections were cut, one, 600 feet south of the footpath, through the bank and ditch, the other, near the south end of the dyke, through the bank only. They are shown in Fig. 4. No evidence as to the date of the earthwork was found in either. For section I a trench 3 feet wide and 40 feet long was cut across the bank and ditch, and was widened to 6 feet in the centre of the bank where small potsherds were found. The soil of this wood " heals " very completely, and the section of the ditch was obscure, but it seems to have had a blunt V shape, the bottom being about 3 feet deeper than at present. When it had silted up to its present level, the ditch was used as a cart track. In the western rut was found a horse shoe, of the small medieval type. The old surface line under the bank was marked by a layer of very clean white sand with some pebbles. Over this was a bank about 14 feet wide and 3 feet 6 inches high of clean whitish sandy gravel with some thin horizontal ironstained lines. Behind this bank was a layer 5 feet wide by 1 foot thick of very hard gravel, stained dark brown. Above this was a layer of loamy sand, dirty yellowish, with many pebbles near the top of the bank, becoming greyer and more sandy over the gravel layer. In the white sand layer marking the old surface line, as indicated, three small fragments of pottery were found. Mr. C. F. C. Hawkes has examined them, but they are so small that they cannot be dated, except to say that they are probably not Roman and not before 100 B.C. Their description is : size 0'2x0-4x0*15 ins. thick, pink surface, grey centre; 0-9x0'5x0-25ins. thick, slightly curved, very sandy grey clay, grey inner surface, brown outside ; and a fragment about 1 in. square x 0 • 3 ins. thick, of soft red clay. The appearance of this section suggested that the original dyke was formed by the bank of clean whitish 20 EARTHWORKS IN JOYDENS WOOD, I I I I I I IS 3T7T-T M ,,**>< al I I _'N*»......... •*,:f-•»...I Z so t<- a a z< U5 -1 <> O «f a n x< > a. > >- 0 z e O CO < in t - X O LO X a O O V) z UJ Q >• O LU >- Q I BEXLEY, KENT. 21 gravel, with the hard gravel layer forming a track behind the bank, and that the dyke had later been enlarged by the addition of the layer of loamy sand. As mentioned above, this part of the earthwork is of stronger profile than is usual. Section I I was dug across the bank only at a point near the south end of the dyke. Its object was principally to determine whether the hard gravel layer found in Section I did in fact extend for the full length of the earthwork. Time did not permit a section to be taken across the ditch. The original surface and subsoil were discoloured sandy gravel, containing a few burnt pebbles. Near the front of the bank was a hollow filled with grey sand, apparently natural. The dyke was formed principally of dirty yellow sandy gravel. In the front of the bank was a soft sandy layer, stained dark brown. It does not seem to have served any structural purpose. At the back of the bank the hard dark brown gravel layer again appeared, 7 feet wide and about 9 inches thick. A few flint flakes were found in the dirty yellow gravel. Section II therefore confirms the conclusions based on Section I. It is certam that the dyke was in existence before the ninth century, that it was interrupted at a steep sided valley, that it was in parts enlarged at an unknown date, and that it was originally followed by a hard gravel path behind the bank. This last feature has not been recorded from any similar work, but it is to be hoped that it may ultimately be found in a dated dyke. It is also of interest as suggesting that the line of the earthwork was intended to be patrolled. The date of the dyke must remain uncertain, but the evidence points to the Dark Ages, rather than to the Roman or pre-Roman periods. Dr. R. E. M. Wheeler has recently drawn attention11 to the dyke systems surrounding London12 and has argued convincingly for a fifth or sixth century date for the Chiltern Grims ditches and the Pinner and Bexley dykes. It has been noted in the papers referred to that these dykes do not usually descend into the valleys, but it does not seem to have been pointed out that if the valleys themselves are 22 EARTHWORKS IN JOYDENS WOOD, regarded as obstacles, the Chiltern dykes and valleys together form a continuous frontier for a considerable distance, and although the Pinner and Bexley dykes are much shorter, the same argument applies to them. This suggests the hypothesis that the dykes belong to a period when the uplands were cleared and the valleys still wooded, rather than that they protected cultivated valleys. This point, however, does not affect the main argument for dating, that the Chiltern uplands with their covering of heavy clay can hardly have been cleared before the end of the Roman period. It may be noted that parish boundaries seem to avoid rather than to follow the dykes, but as it is certain that the earthworks are not later than the Dark Ages, this remains equally remarkable whatever dating is proposed. Possibly the boundaries follow the actual edge of the original clearing. Offa's and Wat's dykes show the same phenomenon. It has been necessary to consider the Chiltern remains in order to show the similarities between them and the Bexley example, but it would be out of place in a paper describing the Bexley earthworks to enter into a general discussion of the purpose of these dykes. The distribution of relics in the Cray and Darenth valleys also points to a date in the Dark Ages. The Iron Age is almost unrepresented, Romano-British remains are equally frequent in both valleys, but objects of the pagan Saxon period are practically confined to the Darenth valley. Their distribution is shown on the map (Fig. 5), and they are listed in Appendix II. To attempt to date the dyke by these arguments is very hazardous, however, and there can be no certainty until more work has been done on similar remains. The Chiltern Grims Ditches offer the most promising field for excavation. If the dyke in fact belongs to the Dark Ages, its construction must surely be connected with the fighting in this district recorded in the Anglo-Saxon Chronicle as the battle of Crecganford, in A.D. 457, and by Nennius as the first of the battles of Vortimer, at the river Darenth,18 but in the present state of our knowledge it would be unprofitable to attempt to assign it to any particular phase in the struggle. BEXLEY, KENT. 23 I am deeply indebted to Mr. M. D. V. Holt, the owner of the western part of the wood, for permission to examine the site, and for much help in locating the earthworks ; to Mr. Hawkes for examining the pottery; to Dr. R. E. M. Wheeler and Mr. 0. G. S. Crawford for much help and advice; and to all those who have either helped me with the survey or permitted me to examine earthworks on their land. The objects found are now in the possession of Mr. Holt. APPENDIX I. BOUNDARIES oir BEXLEY, A.D. 814u (see Map, Fig. 5). " iErest up of crsegean on Man rithe " (First up from Cray to foul brook). The " foul brook " is a small stream which runs almost straight from Stankey Wood towards Crayford railway station. " andlang rithe oth thone fsestendic " (along brook to the strong dyke). The parish boundary follows the dyke from Dartford Lane as far as the footpath through the wood, except for a slight deviation near the path. " andlang dices oth thaet gebyhte " (along dyke to the bend). The bend is the point on the footpath at which the parish boundary leaves the dyke, which here turns sharply through a right angle towards the south. " of tham gebyhte and lang hagan oth cyninges healh. thanon andlang hagan ut on crsegean " (from the bend along hedge to king's hollow, thence along hedge out on to Cray). Cyninges healh must have been near Cavey's Spring. " andlanf crsegean oth thone hagan. andlang hagan oth psethfeld. thanon andlang hagan oth aescburnan (along Cray to the hedge, along hedge to pathfield, thence along hedge to ash stream). This defines the south side of the parish. " iEscburnan " must refer to the stream at the south-west corner of the parish. " of tham burnan andlang hagan on casincgstrset " (from the stream along hedge on to " casincgstrset "). As pointed out in E, " casincgstrset " refers to the Roman road now known as the Watling Street. 24 EARTHWORKS IN JOYDENS WOOD, tIA - LYTLANLEA. ENEDE LEBROO f„;-» IOW> * •-*/§£ 1 (t.n I .™ i » E fc/"*5 <3ss&™ " \ SMS SCALE 5 MILES WOODED CLAYS TRACKWAY LIGHTLY WOODED GRAVELS OPEN CHALK ALLUVIUM DY KE ROMAN ROAD • SAXON CEMETERIES * MISC.SAXON OBJECTS » » » » x BEXLEY PARISH BOUNDARY Ero. 5. Saxon remains in the Cray and Darenth valleys. BEXLEY, KENT. 25 " east andlang strsete on scoffoces sse. thanon north andlang strsete oth lytlanlea " (east along street to ?, thence north along street to little clearing). This part of the survey follows the boundary of East Wickham parish as shown on the 1898 edition of the 6" O.S. map (Kent II. SE.). It is not shown on the latest edition. The second street runs north from Bexleyheath towards Abbey Wood, along the line of an existing track. " Lytlanlea " survives in the name of the Hundred of Little and Lessness. " thanon east andlang mearce oth enede mere suth rihte of tham mere to burnes stede " (thence east along balk to duck pool straight south from the pool to stream place) " Bumes stede " survives as Bursted Wood. " thanon andlang hagan oth casincgstraet. andlang strsete oth thone calewan telgan. thanon ut on crsegean. swa eft on Man rithe " (thence along hedge to Casingstreet, along street to the bare boughs, thence out on to Cray, so again to foul brook). The " bare boughs " must have been somewhere near Gibbet Hill, and may perhaps refer to the gibbet. The text and a translation of the survey are given in E, but no points are identified except " casincgstrset ". The identification of the " fsestendic " is left in doubt, but it is suggested that it may be the track AB. This identification would be just possible, though improbable, if no other examples were known of the type of earthwork to which this name was applied, but three other " fsestendics " can be identified, and in each case the reference is to an earthwork similar in size to the dyke in Joydens Wood. The name is applied to two sections of the Chiltern Grim's Ditch,15 to a dyke on Hartford Bridge Flats near Camberley,16 and to the Devil's Ditch near Andover.17 The exact meaning of the word is uncertain. The most usual translation is " fortress-dyke," but Mr. 0 . G. S. Crawford sugests that " strong dyke " is the correct meaning. This suggestion is supported by the absence of anything that could be called a " fortress " from the neighbourhood of the dykes, except in Joydens Wood, where there is very strong reason to regard the adjacent earthworks as more recent. But whatever the true interpretation, the name surely implies that the Saxons regarded these works as of some military value, in spite of the tendency of modern archaeologists to dismiss them as mere boundary marks. 26 EARTHWORKS IN JOYDENS WOOD, APPENDIX II. LIST OE OBJECTS PBOBABLY OE THE PAGAN SAXON PBBIOD IN THE VALLEYS OE THE CBAY AND DABENTH. (See Map, Fig. 5.) 1. Iron Spearhead, found in the peaty mud of the river Cray, High Street, Old Bexley. (Dartford Museum.) 2. " Ornaments" and 3. " miscellaneous." Now lost. Said to have come from graves. Arch. Cant., XVIII, pp. 307,316. 4. Cemetery, Lullingstone. VCH Kent, I, p. 378 (q.v. for other refs.). It seems probable that the Eynesford cemetery (VCH, I, p. 385), which is said to have produced a bronze bowl similar to that found at Lullingstone, is the same site, the confusion being due to its position midway between the two villages. 5. Pot and spearhead, probably with burial, from Charton Manor, Farningham. (Dartford Museum.) 6. Cemetery, Horton Kirby. (Dartford Museum.) 7. Cemetery, Horton Kirby (on either side of railway). VCH, I, p. 377. 8. Burials at the Powder Mills, Darenth. A.C., XVIII, pp. 307, 314. Spearhead and shield-boss in Dartford Museum. 9. Spearhead from Dartford Brent. (Dartford Museum.) Tumuli are recorded here by Spurrell, A.C., XVIII, p. 307. 10. Burials, Littlebrook. A.C., XVIII, pp. 307, 316. 11. Littlebrook WaUs. A.C., XVIII, pp. 307, 311. Although none of these, except perhaps No. 4, can be assigned to the date suggested for the dyke, the distribution shows a great difference in the intensity of occupation in the two valleys during the succeeding century and a half. Of the three sites in the Cray valley, No. 1 may well be the relic of a wandering hunter, and Nos. 2 and 3 are both east of the natural scarp which continues the line of the dyke to the Cray. I t must be emphasized that this difference is not due to a variation in the intensity of archseological fieldwork. About twenty Roman sites are known in the Cray valley, and the " development " of the last few years has been carefully watched by Mr. Greenfield, but no Saxon objects have been found. Littlebrook (No. 11) is mentioned in pre-Norman documents, and its inclusion here is perhaps justified by the presence of the burials near by. A note on its probable archseological importance appears in the Miscellaneous Notes in this volume. BEXLEY, KENT. 27 The writer is indebted to Mr. Norman Cook for information as to the sites recorded at the Maidstone Museum; to Mr. E. Greenfield for details of his work in the Cray valley ; and to Mr. S. Priest for permission to mention unpublished material in the Dartford Museum. REFERENCES. I F. C. J. Spurrell, "Deneholes and other caves with vertical entrances." Arch. Journ., XXXVIII, pi. I, p . 404. On June 24th, 1882, t h e Geologists' Association visited the Deneholes of Bexley under the leadership of E. C. J. Spurrell and T. V. Holmes. Proc, G.A. Vol. VTI, pp. 400-3. a F. C. Elliston Erwood, " The Roman Antiquities of N.W. Kent," Journ. Brit. Arch. Ass., N.S. XXXIV (1928), p. 165. 3 Col. O. E. Ruck, " Archseologia Militaria, Joydens Redoubt, Kent," The Boyal Engineers' Journal, vol. IV, No. 1 (July, 1906), p . 16. 4 VCH Kent, vol. I , p. 404 ; vol. I l l , p. 146 ; Journ. Boman Studies, XV, p. 245. The occasional references of local historians add nothing to our knowledge of the site. 6 One set at the Ordnance Survey Office, Southampton, the other in the possession of Mr. M. D. V. Holt. Most of the earthworks can be traced without difficulty on these photographs. 6 N. Cook and R. F. Jessup, " Excavations in Rose Wood, Ightham," Arch. Cant., XLV (1933), p . 162. ' F. C. J . Spurrell, " Dartford Antiquities," Arch. Cant., XVIII (1889), p. 307. 8 Surrey Place Names ((EPNS,) vol. XI, 1934), p . 71. 9 A. W. Clapham, Lesnes Abbey (Cassio Press, London, 1915), pp. 7 and 22. The following documents are referred to : Charter to Strood Hospital A.D. 1194. Thorpe, Reg. Roff., p. 641. Statement of Possessions, A.D. 1472. P.R.O. Rentals and Surveys, G.S. Roll 357. 10 Antiquity, VIII (1934), pp. 218-222. I I R. E. M. Wheeler, " London and the Grims Ditches," Ant. Journ., XIV (1934), p. 254. 11 M. W. Hughes, " Grimsditch and Cuthwulfs Expedition to the Chilterns in A.D. 571," Antiquity, V (1931), p . 291. For the topography of the Chiltern dykes: O. G. S. Crawford, " The Chiltern Grims Ditches," Antiquity, V (1931), p. 161 ; and O.S. Map, 1 inch to 1 mile. Fifth (Relief) Edition, Sheet 106. 18 "Historia Nennii. Cap. XLVII," Mon. Hist. Brittanica (1848), p. 69. 14 W. de G. Birch, Gartulariwm Saxoniaum, voL'I, No. 346. Grant by Coenwulf, king of the Mercians, to TJulfred, archbishop of Canterbury, of land at Bexley, A.D. 814. The text is taken from a nearly contemporary copy of the charter. 16 Antiquity, V, pp. 294-6. 16 A. H. A. Hogg," Dyke on Hartford Bridge F l a t s , " Proc. Hants F.C., vol. X n i , pt. I, pp. 70-74. 17 " Perambulation of Fynkely, A.D. 1323," O. G. S. Crawford, The Andover District (Oxford, 1922), p . 84.